danbickell

The VF-1A in that screengrab always seemed like a mistake to me. If you look at it closely, frame by frame, it is exactly a Kakizaki skull squadron scheme, only with parts colored with tan CF style. That's why it has the black stripes on the arms (and legs), the arrows on the head, and it even has the green and black chest pattern, as well as the matching squadron roundel on the FAST packs. All green/black. Maybe it was cut footage of Kakizaki from the beginning of the movie, and they decided to use it for the battle outside the window later (after Kakizaki was dead) and re-colored it to be a CF. A better DYRL CF would be like one of the ones shown in the hangar, with red circles on the legs, and different squadron roundels for the FAST packs. This offering from Yamato just looks like the same thing as the TV CF, but with the DYRL A head, maybe the DYRL gray instead of white, and I would assume DYRL pilot and fixed-posed hands.

I picked up the 3 DYRL VF-1As. Shipping to Los Angeles came out to $60 via EMS.

Wow, I completely forgot about that until just now! Unfortunately, no, I don't seem to have any of the A models, or most of the texture sets any longer. I lost all of that stuff in a hard drive crash, except some copies of the VF-1S that I had converted to test out at work. Doing a quick google search, I see there are still some images left from that though: That model serves as a good example of trying to do too much with too little. It would look ok at a distance though, at the low resolutions of the day, and that was the whole trick back then. Bonus! There are a few images of the Quake 2 model in there too! 900 polys and a single 8-bit 256x256 texture, circa 15 years ago... Yeah, I don't think they ever thought that out. Even funnier is that with the original landing gear design the main actuator is directly behind the gear strut (rather than offset to the side like I have it, because it won't actually fit any other way), so the A head gun needs to be angled way down to the entire time the gear is down in order to clear the actuator. At least with the offset actuator, the gun just needs to pivot to clear the doors and wheels, and then can pivot back into place.

Ok, I got one of those old animation test renders put together as a video: http://s362.photobucket.com/albums/oo68/danbickell/?action=view&current=nose_gear_anim_test4_6.mp4 This is from about a year ago, so there was no detail on the exterior of the nose yet. I was just spending forever getting the landing gear to work right. Bummer, I thought I had a version somewhere after the head was in there, with the head gun moving to get out the way too.

Doing just a flight sim for the fighter is one thing, but animating battroids and gerwalks is another can of worms. In my experience, pretty much everything these days uses skeletal animation in games. In some cases it might not seem so, and might be transparent to the artist, but it is still getting "rigged" for skeletal animation during export or in the engine itself. Some engines just look at the pivot points of each object, create a "bone" there with rigid "skin", and link it to the next bone via the object hierarchy. The end product is the same thing, and it generally works for any rigid-bound hard-surface stuff. One way or the other, I'm really just looking at it from a performance standpoint. Dreaming about a pie-in-the-sky ultimate Macross sim/game, designed for current-gen consoles and/or PC, how would I go about putting together a real-time game asset out of this VF-1? If you've got a world to render, inhabited with a decent number of mecha, poly count and bone count end up being the major limiting factors involved. No biggie. I'm thinking that a 40k poly VF-1 requiring 60+ bones could work really well, as long as the rest of the game design is kept manageable. Current-gen consoles could easily struggle if you tried to render too many of those at once. I can guarrantee that PS3 game that is coming with the DYRL BluRay is lower poly and uses less bones, but why would I aim to replicate that? I hear you about the landing gear animations. I had the nose gear all rigged up and rendered out some animation tests in Maya last year. Like I said, there was a good 20 bones just for the gear, and setting up constraints and timing it all took quite some effort. I'll have to see if I can put together an AVI or a GIF from the renders...

Happens to the best of us! I'd love to go back in time and tell that guy that he would still be modelling the same old subject matter in 2012, only with HUNDREDS of times the polycount. Maybe in 2025, we'll be looking back at the quaint old-school 2 million poly VF-1 model and be amused at how so much was done with so little, while working on the multi-billion poly holographic simulation model!

Yeah, most of my animation learning came from rigging and animating my various VF-1 models back then, though I haven't done any animation professionally since the 90s. Back when skeletal animation was the new thing, I learned Lightwave just so I could try it out with a transforming rig. The Quake 2 stuff was an interesting challenge, since the format consisted of a bunch (something like 200) of key frames (so it was just 200 copies of the same geo, in different poses), and interpolated between them. So, I just did "real" skeletal animations and strategically picked frames to save out so that everything would work half way decently. A lot of people contacted me about that back then, wondering how it was even possible given the limitations. It was all a matter of tricky timing, so that the interpolation was mostly confined to just the linear motions. Later, as skeletal animation became the norm in games, the skeletons we had were very basic in the early days, and usually limited to a one-size-fits-all scenario. Best I could do there was to rig a battroid model to the skeleton as best as possible, and have a laugh at a human-sized VF-1 battroid running around in the game (with parts clipping all over the place), driving cars and kung-fu fighting and such. A few times, I've set up VF-1 fighter models as cars (hovering above invisible wheels) and drove them around (set to be faster than any car in the game, of course!). Most everything I've worked in on recent years has been all mo-cap, and likewise limited to a specific skeletal rig with around 20-30 bones. I will neither confirm or deny that any valkyries ever inhabited the Call of Duty universe, despite the freedoms that existed for vehicle rigs and the eventual increase of the bone count limit that was made for a particularly complicated game asset I worked on. Needless to say, a modern VF-1 game model that fully transforms and has all the bells and whistles would require a fairly complex rig. The nose landing gear alone currently needs about 20 bones, with all the little hydraulic actuators for the doors and such. The canopy currently uses 5. A lot of the little things would need to be simplified, deleted, or just not animated to make it reasonable for a game. I'm thinking something in the ballpark of 100 bones would be good enough for a good sim-type valkyrie game. Something more arcade-like could maybe get away with <60. Did you guys happen to see the screens from the new PS3 game that comes with the DYRL BluRay set? http://www.ncsx.com/2012/040212/ncs0402u.htm That is looking pretty good so far, and certainly better than previous games. I'm 100% confident that I could put something together that would easily rival it, though. Of course, that's easy to say when you can spend all the time you want doing it, and don't have any tech limitations of a certain engine framework to work within. And THAT is precisely what I love about working on this VF-1 model. Then again, that's exactly why it will end up taking so long to ever finish!

Thanks so much! Looks like I might just have some time this week to get back into it. That has actually been in the plans all along. I've been trying hard to not get side-tracked into starting to do it with what I have already, just to see how it turns out. Making efficient game models seems to be a dying art these days (since you can throw so much geometry and textures at modern game hardware), but that's how I've made my living for 17+ years, and it is naturally what I'm best at. I have a long history of putting together versions of my VF-1 models tailored for whatever game engine we were using at work at the time. It's been years now since I've had a modern (good) model to try, and we currently have some interesting shaders I would love to try out on it. Of course, I've always had to keep these things private, and I actually have got in a bit of trouble for this sort of thing in the past (ie. reps from publisher are visiting and get a glimpse of the "giant robot" in my test level, legal guy freaks out, and producer is all over my bosses about why I supposedly don't have bandwidth to do what they've been requesting... oops!). Luckily (I think), most of that stuff isn't around for anybody to see anymore! The Quake 2 player models I did still seem to be floating around a few places, but I couldn't find any decent sized images of them (whew!). Those were 900 triangles with tiny 8-bit textures, and transformed to all 3 modes (battroid for standing animations, gerwalk for crouching animations, and fighter for the taunt). That was pushing the limits back then, but totally out-classed by the PS2 and PSP games that came since. I did manage to dig up this little blast from my past, circa 1999 (try not to laugh too much!): That was a 4000 triangle model, fully transformable, with pilot, landing gear, and even movable wing control surfaces,flaps and airbrake. This was also some experimentation in the early days of bump maps. That was ahead of the times (PS1 days), and the PS2 game models that came in the years to follow left me a little disappointed. These days, it would be a minimum of about 20k triangles to do it decently, with diffuse, spec, normal, and gloss maps. I would probably aim closer to 40k, though 100k+ models aren't out of the question at all anymore. It just depends on what else you have to render. So far, this model only clocks in around 650k triangles. It is built more efficiently than it might look (old habits die hard...). The whole thing might come out to be in the ballpark of 2 million. Even that wouldn't be a problem for real-time on high-end PC gaming rigs, under the right circumstances. In any event, I am very much interested in making it all available to the talented folks in the community who would do cool stuff with it. I'd love to see people make movies and games and such with it! Now I just have to get it finished...

Yeah, I got the same from CD Japan. I tried the "special request" option. I ended up ordering the LE through YesAsia instead.

For anybody who couldn't right-click-save-as:

I've spent plenty of time looking at that drawing in the gold book over the years, but thanks for reposting it here (and thanks for the other gold book images you had on your site that I used in my previous post!). That does look similar to an ejection seat rail, but at closer inspection, it can't really work that way. The angle only matches the botton rear corner of the seat, not the majority of that groove up the back of the seat which is at a different angle (which allows the space for it to tilt back for the battroid position). What that part does happen to line up with exactly is the floor of the cockpit behind the seat (which is the roof of the gear bay, the deeper part where the wheels fit). So, I built that portion of the cockpit floor behind the seat with a matching groove to fit that portion of the seat frame. What would work is having that bottom angle of the grooved frame on the back of the seat be the connection point for a hinge mechanism that tilts the seat back for battroid. That would have to exist in that groove it sits in in that angled section of the cockpit floor. No biggie, but I just haven't modeled anything like that in there yet. If there were any hydraulics involved, the space limitations would require that they be taking up space inside the landing gear bay, in the space just above the tires, working around the hinge for the main gear actuator (which is offset to the side in my version anyway). That part, at least, is doable. Still, it doesn't put the seat in a position that quite matches where the drawings show it relative to the front console. It would still require some fudging and/or a moving front console and that whole can of worms. Maybe I will eventually work something out that will work. Believe me, I would really love to!

Yeah, that Master File battroid cockpit is certainly interesting, I'll say that... The only point to the whole landing gear bay moving seems to be for the seat to raise out behind the head. They also seem to have gone with the idea of the screens being on the undersides of the front console and the rear box, which both move quite a bit to end up in front of the pilot in battroid. (I guess that is another reason to move the gear bay, so the WHOLE front console can move up in front of the pilot!) It isn't a horrible idea, but it doesn't remotely fit with anything in the original designs. The screens on the inside of the canopy shield makes a fair amount of sense, but again, doesn't match up with the designs at all. The DYRL battroid cockpit screens are individual consoles that are close to the pilot, with controls mounted to them. Another problem with this idea is that the shield only really covers the front portion of the canopy, and the rear portion is covered up by the chest piece. The inside of the shield should really only be visible at the bottom half of the battroid cockpit. If we were to assume that the shield was long enough to cover the entire canopy (which was shown that one time in fighter mode in the TV series...), then there are big problems with it fitting anywhere inside the chest and back in fighter, and certainly with clearance for it during transformation (which is why the toys are all done the way they are). Of course, the DYRL canopy itself acts as a 360 degree HUD, so I don't see why whatever system for that (projection or internal LED display) wouldn't work as view screens anyway. This seems like the most logical idea, and it kinda fits as a predecessor to the future holographic displays. It seems a lot more thought went into the battroid cockpits of the later VF designs, and given all the problems there with the VF-1, I can see why!

Not sure what you mean... Proportionally, the nose will look much like the Yamato 1/60 in battroid (which I think is a good thing!). I'm using proportions from the Hasegawa 1/48, which matches up very closely with the Yamato 1/60 (at least on the top side, which is what is visible in battroid). I'm guessing you mean the cockpit parts. I've played around with it, but ultimately decided not to try to make it reconfigure for battroid. Although the DYRL battroid cockpit certainly works better than the TV version, it still is quite problematic. I built the seat in it's entirety, with provisions to angle the head rest. I did not build any mechanism to tilt it though, and I'm not sure where it would fit (there is no room, the floor of the cockpit is practically the roof of the nose gear bay). I've played around with it, and the seat does fit with the proper angle and everything (which really isn't all that much of angle, as opposed to the 90 degree rotation on the TV version). There certainly doesn't seem like there's enough room for all the stuff that has to maneuver around it, though. Here's some old renders of the seat by itself: The arm rests would need the underside frame built and the hydraulics to angle them up with the seat (plenty of roof for that, at least). Likewise, the stick and throttle would need the mechanism to extend them out of the side of the console and up into their battroid positions. I did build the throttle to rotate to the vertical position, at least. The various screens are the biggest problem. The front screen needs to come out of the front console (the top of the console is visible at the bottom of the forward-looking picture). The whole top of the console would have to open up for this, as the back of the front console doesn't match anything on the console. It would be much easier if the top of the console WAS the back of the front screen, but it just doesn't match at all. Also, the mirrors need to fold back somehow, without being in the way of the screen parts coming out of the top of the console. The rest of the screens all need to fold out of the "box" behind the seat, somehow. Again, it would be great if that box folded apart and extended, with all the screens on the inside, but none of the back surfaces of those screens remotely matches anything on the box. All the extra battroid controls need to be attached to those screens as well. There's actually quite a bit that would have to magically fit inside that box. And then, if you wanted to go so far as to have the seat be able to raise up for battroid egress/ingress (with the head swiveled forward, another problem altogether in itself), everything to do with the big vents at the back of the cockpit (and whatever ducting is behind it, presumably leading to the exterior vents) needs to magically go somewhere to get out of the way of all that. So, basically there just is no way to do it and keep the detail anything close to accurate to the fighter cockpit drawings. I'd rather have the accurate fighter cockpit. Maybe I'll do separate battroid cockpit model, accurate to the lineart, just to do it. I'd rather have two separate-but-accurate models than a severely accuracy-compromised transformable model. If I had those 2 models, only then might I try to make up a transformable model that would work, but wouldn't match the line art.

The gear doors do look pretty slim in that most recent render. Part of it is that I don't have the red portion extending as far as that F-14 (the contrast makes a big difference visually). With the limited space in there (it is REALLY tight), I decided to stick closer to more modern aircraft. If you look at F-18s, the doors are quite a bit slimmer compared to the F-14, with smaller hinges too (and red only on the outer lip). Here's a few of the older renders that show it more clearly: Off the top of my head, the doors are at well over an inch in total thickness, with maybe 5 or 6 mm of outer skin (I'll measure when I get home). That seems about right, but I kinda agree that it might look better thicker. I could probably make it fit with just an extra set of indentations for tire clearance. Right now, they nearly touch, and I wouldn't want to shrink those any (they are a bit on the small size, and were always drawn bigger than could possibly fit). Good critique! I'll see what I can do...

I have, and enjoy, all the Master Files books. There is some great work in there, but I actually don't like most of the liberties they took with the designs. They did some pretty wacky stuff to get things to work, and it often comes at the cost of changing the designs too much just to get one little feature shown in one episode to kinda work. In that respect, I've found it more valuable to simply see an idea already explored so I can more easily reject that approach. I'm also not a fan at all of the proportions of the model they've used for all their diagrams. It looks decent in fighter, but not so hot in battroid or gerwalk. The real value of these books, for me, has been motivational. I see a lot of effort and good work put into exploring the details of this stuff, but it ultimately leaves me unsatisfied, which motivates me to try to do it better. He actually just recently found me on Facebook and friended me. Haven't chatted with him much yet, though. He's still in game development, and probably just as busy I am.

Thanks guys! I've been working on the chest and LERX this week, trying to get everything to work. The fairings on the underside that cover the top portions of the intake and thighs are problematic for transformation. I guess that is why they have been under-sized or left out on transforming toys and models. I've got a version worked out where the front portion of the fairing (part of the LERX) fits inside the middle portion (part of the chest) in battroid mode, but they have to clip through each other to get there. It would look fine in either mode, but not mid-transformation. I've been hung up on that, but I might try adding a hinge point to the LERX fairing to get it to work. Leaving any room for the mechanism to swing the legs down to the nose is problematic too. Not to mention room for the battroid side covers... ugg... Yeah, I went ahead and put reflection materials on the mirrors, and on all the glass as well. The interesting thing about the mirrors is that they would need to be angled quite differently than they have always been drawn in order to actually get functional rear views outside the cockpit. Either that, or they would have to be some kind of fish-eye mirrors. I played around with it to get a functional view, but that angle looks wonky from the outside. I would hugely appreciate that! 300 DPI would be fine, but I certainly wouldn't turn down 600 DPI. I'd love scans of the decal sheet too, please.

Thanks guys! Believe it or not, I finally started making progress on the rest of the model over the weekend. Here's a quick preview: Still based heavily on the Hasegawa 1/48, but tweaked with DYRL differences. More of that will be apparent once the vents are detailed. I had to go ahead and cut out the airbrake panel early to square away the space inside. The moving parts for the head base are directly below it in fighter mode, so I had to tweak some of that to fit. There is a decent amount of room in there, but I will have to be careful with it to leave enough space for the heat shield to fit in between somehow. Fun times ahead!

Thanks, but there certainly needs to be textures for all the markings and paint schemes and weathering and such. An untextured CG model is like an unpainted model kit. Texture is what will really bring it to life, with all the squadron markings and little "NO STEP"s and such, and subtle faded and chipped paint, and grime. That will be the fun part (though UV mapping it all, not so much).

Well, my girlfriend went to Mexico for the weekend with her parents, so I had the house all to myself. I ended up pulling an all-nighter getting the model I imported into Max all cleaned up and presentable. I never remember having this much trouble transferring models between different formats between Maya and Max, but that's always been game assets that are much less compicated. The .fbx exports came through with the hierarchy, rigging, and constraints all intact, but way too much corrupted geometry (I tried several format versions and options). I ended up using an .obj export that came through with fairly clean geometry, but lost all the hierarchy and pivot points and such. I ended up re-triangulating and re-smoothing just about everything (currently at 650k polys), and consolidated the mess of separate materials with new fancy Mental Ray Arch & Design materials, ready to plug-in textures once I get that far. Finally, I've got it all sorted, short of re-rigging the landing gear and canopy for animation (though I sorted out the pivot points and hierarchy to get everything to work again manually). I'm quite happy with the results in Max, and now I can start moving forward with this again! Of course, I saved out plenty of renders along the way, and put together a selection of them for you guys today. I kept them at full-resolution 1920x1080 (except for a couple crops) this time around. Enjoy! Thanks to you all for being so supportive and enthusiastic. I truly appreciate it.